200900981 九、發明說明： 【發明所屬二 發明領域 本發明係有關一種可於以有限元命、、表w i去進行變形解析 時，選出破壞危險部位之破壞預測方法、演算處理^ 程式及記錄媒體。 ~ 10 15200900981 IX. OBJECTS OF THE INVENTION: [Technical Field] The present invention relates to a damage prediction method, a calculation processing program, and a recording medium for selecting a damage-destroying part when performing deformation analysis using a finite element life and a table w i . ~ 10 15
I 20 近年，汽車業界已將可降低車搞時對乘客之傷 體構造之開發視為當前之要務。又，同時使車體輕量化以 7低燃料成本亦甚為重要。為解決該等問題，已就更高強 紅材料’尤錢鋼㈣檢討高強度鋼板之適祕。缺而， ::強度之提昇將導致成形性之劣化，且為擴大適 。成雜，尤以改善伸緣成频為㈣。 發。㈣題，伸緣絲性較佳之材料已進行開 專利文獻1即揭露麵控制肥粒鐵及變勒鐵 m路/規定塑性各向異性與特定杨之拉伸測試之 ^展^提昇伸緣絲性之㉟合金板。 之零件之!形可否並非僅由材料特性決定 因此，為發/祕'間'月條件、成形條件等複雜之影響。 此為發揮優良之材料特性，， 該等複雜條項與材料共同適當設定 土、述目的，而適用了數值解析技術。 揭露有使时限元素法而預測成形時之成 5 200900981 形瑕疵之破壞及皺褶之方法。依此，可使用有限元素法進 行解析，並利用所著重之元素之應變及應力資料而進行破 壞及皺褶之發生判定。然而，使用上述方法時’必須對應 解析對象而依適當大小進行元素分割’若元素分割不適當 5而進行解析，則預測結果可能過度或不足而無法對應實際 情況。 【專利文獻1】特開2002-60898號公報 【專利文獻2】特開2006-257506號公報 【專利文獻3】特開平8-339396號公報 10 【發明内容】 發明揭示 如上所述，使用有限元素法而預測成形時之成形瑕麻 之破壞及皺褶之發生時，藉習知技術確實選出破壞危險部 位仍極為困難。 15 本發明即有鑑於上述問題而設計者，其目的在提供一 種可於藉有限元素法預測破壞時，輕易且確實地選出破壞 危險部位之破壞預測方法及裝置、演算處理裝置、程式及 記錄媒體。 本發明人著眼於變形集中在破壞危險部位，且其周圍 2〇產生較大之變形梯度，而檢討破壞預測方法，並發現可確 實判別破壞危險部位。本發明之要旨如下。 1. 一種破壞預測方法，包含有：第1步驟，使用有限元 素法，以第1領域及大於前述第丨領域之第2領域分別分割解 析對象零件，再進行成形解析；第2步驟，就前述各第崎 6 200900981 域及别述各第2領域分別算出最大 = 及’第3步驟’在相當於前述解析對=1¾板厚減少率； 置上，選出前$最大主應變或前述板」牛之同-部位之位 域與前述第2領域之差值大於預定值：？率之前述第U! 前述解析對象零件之破壞危險部位，34第1領域’作為 2.如1·之破壞預測方法，前述第丨步驟 析對象零件之_之_，決定前述& ’依與前述解 第2領域之大小。 或之大小及前述I 20 In recent years, the automotive industry has considered the development of the injury structure of passengers as a current priority. At the same time, it is also important to reduce the weight of the vehicle at the same time. In order to solve these problems, the high-strength steel plate has been reviewed for the higher strength of the red material ‘Yu Qiangang (4). Lack of, :: the increase in strength will lead to the deterioration of formability, and to expand the appropriate. It is complicated, especially to improve the frequency of extension (4). hair. (4) The material with better edge-stretching properties has been opened. Patent Document 1 discloses the surface control of the ferrite and the iron-iron m road/specified plastic anisotropy and the tensile test of the specific Yang. 35 alloy plate. The shape of the part can be determined not only by the material properties, but also for the complicated influence of the 'month' condition and forming conditions. In order to exert excellent material properties, these complex items and materials are used together to appropriately set the soil and the purpose, and numerical analysis techniques are applied. It is revealed that there is a time-bound element method to predict the formation of the shape. 5 200900981 The method of damage and wrinkles of the shape. Accordingly, the finite element method can be used for analysis, and the strain and stress data of the heavy elements can be used to determine the occurrence of damage and wrinkles. However, when the above method is used, it is necessary to perform element division according to an appropriate size in accordance with the analysis target. If the element is not properly divided and analyzed, the prediction result may be excessive or insufficient to correspond to the actual situation. [Patent Document 1] Japanese Laid-Open Patent Publication No. Hei. No. Hei. No. Hei. No. Hei. No. Hei. No. Hei. It is extremely difficult to determine the damage to the dangerous part by the conventional technique when predicting the occurrence of the formation of ramie and the occurrence of wrinkles during forming. The present invention has been made in view of the above problems, and an object thereof is to provide a damage prediction method and apparatus, a calculation processing apparatus, a program, and a recording medium which can easily and surely select a damage-damaged part when predicting damage by a finite element method. . The present inventors focused on the fact that the deformation focused on destroying the dangerous part, and a large deformation gradient was generated around the ridge, and the damage prediction method was reviewed, and it was found that the damage dangerous portion was confirmed. The gist of the present invention is as follows. A method for predicting damage, comprising: a first step of dividing a part to be analyzed in a first field and a second field larger than the first field by a finite element method, and performing forming analysis; and the second step Each of the Dizaki 6 200900981 domain and each of the second fields described separately calculate the maximum = and 'the third step' in the corresponding analysis pair = 13⁄4 plate thickness reduction rate; above, select the top $maximum main strain or the aforementioned plate" The difference between the bit field of the same-part and the aforementioned second field is greater than a predetermined value: In the above-mentioned U! The destruction target part of the analysis target part, 34 first field 'as the damage prediction method of 2., the above-mentioned third step analysis object parts, determines the above & The size of the second field is solved as described above. Or size and the foregoing
1010
20 •.犹厶之破壞……々欲，珂述第3呆 出大於前述預定值之前述第1領域，則在前^若未選 迷第2領域中至少將前述第】領域設定成更:，弟1領域及前 对述第1步驟、前述第2步驟及前述第3步顿。再依序執仃 前述第1步驟中， 迷解析對象零件 4.如1.〜3.中任一者之破壞預測方法， 以前述第1領域及前述第2領域分別分割前 之端部，再進行成形解析。 个▲ 7哪，使用有限元 素法，將㈣對轉件分割錢數領域，再進行成形解析； 第2步驟’就前述各領域分別算出最大主應變或板厚減少 率；第3步驟’結合隣接之2個以上之前述領域，算出已結 合之前述領域之前述最大主應變或前述板厚減少率；及， 第4步驟，選出前述領域之結合前後之前述最大主應變或板 厚減少率之差值大於預定值之前述領域，作為前述解析對 象零件之破壞危險部位。 6·如5.之破壞預測方法，前述第㊇驟中，R前述領域 7 200900981 分割前述解析對象零件之端部，再進行成形解析。 7. -種運算處理裝置，使用於解析對象零件之破壞預 測方法’包含有11機構’使用有限元素法，以第丄領域 5 10 15 及大於前述第1領域之第2領域分別分割解析對象零件，'再 進仃成形解析；第2機構，就前述各第丨領域及前述各第2領 域分別算出最大主應變或板厚減少率；及，第3機構，在相 當於前述解析對象零件之同-部位之位置上，選出前述最 大主應變或前述板厚減少率之前述第i領域與前述第2領域 之差值大於預定值之前述第！領域，作為前述解析對象零件 之破壞危險部位。 8. 如7_之運算處理裝£，前述第！機構係依與前述解 對象零件之咕之關係決定前述第1領域之大小及前述第2 領域之大小。 2 9. 冑運算處理裝置，使料解析對象零件之破壞預 愛方法包3有.第！機構，使用有限元素法，將解析辦象 ^牛分割《數領域，再進行成形解析；％機構，就前塊 領域分別异出最大主應變或板厚減少率；幻機構，結八 ^之2個以上切述賴，算出已結合之前述領域之 :大主應變或前述板厚減少率;及，第4機構，選出前述， ^結合d之前述最大主應變或板厚減少率之差值大於 位疋值之IT相域，作為前料析對㈣件之破壞危險部 10.種式，可使電腦執行下列步驟：第^步驟 用有限元素法’以第J領域及大於乂 罘貝A及大於則述第1領域之第2領域分 20 200900981 別分割解析對象零件，再進行成形解析；第2步驟，就前述 各第1領域及前述第2領域分別算出最大主應變或板厚減少 率；及，第3步驟，在相當於前述解析對象零件之同一部位 之位置上，選出前述最大主應變或前述板厚減少率之前述 ' 5 第1領域與前述第2領域之差值大於預定值之前述第1領 域，作為前述解析對象零件之破壞危險部位。 11.如10.之程式，前述第1步驟中，依與前述解析對象 零件之η值之關係，決定前述第1領域之大小及前述第2領域 f 之大小。 \ 10 12.如10.或11.之程式，前述第3步驟中，若未選出大於 前述預定值之前述第1領域，則在前述第1領域及前述第2領 域中至少將前述第1領域設定成更小，再依序執行前述第1 步驟、前述第2步驟及前述第3步驟。 13.如10.〜12.中任一者之程式，前述第1步驟中，以前 15 述第1領域及前述第2領域分別分割前述解析對象零件之端 部，再進行成形解析。 I 14. 一種程式，可使電腦執行下列步驟：第1步驟，使 用有限元素法，將解析對象零件分割成複數領域，再進行 成形解析；第2步驟，就前述各領域分別算出最大主應變或 20 板厚減少率；第3步驟，結合隣接之2個以上之前述領域， 算出已結合之前述領域之前述最大主應變或前述板厚減少 - 率；及，第4步驟，選出前述領域之結合前後之前述最大主 應變或板厚減少率之差值大於預定值之前述領域，作為前 述解析對象零件之破壞危險部位。 9 200900981 15. 如14.之程式，前述第1步驟中，以前述領域分割前 述解析對象零件之端部，再進行成形解析。 16. —種記錄媒體，可由電腦進行讀取，且記錄有 10.〜15.中任一者之程式。 5 依據本發明而預測加工零件之破壞，即可降低對解析 條件選擇之依賴性，並輕易且確實地選出破壞危險部位。 藉此，除可降低開發所需之成本，並可對加工零件適用強 度更高之材料，而實現輕量化。 圖式簡單說明 10 第1圖係本發明之破壞預測方法（裝置）之流程圖。 第2圖係顯示粗疏及細密的元素分割之上下限決定時 之模擬結果之特性圖。 第3圖係顯示粗疏的元素分割之上下限決定時之模擬 結果之特性圖。 15 第4圖係顯示細密的元素分割之上下限決定時之模擬 結果之特性圖。 第5圖係本發明之破壞預測方法（裝置）之流程圖。 第6圖係成形測試所使用之素板形狀之說明圖。 第7A圖係顯示無凸緣成形測試之開始前之縱截面之模 20 式圖。 第7 B圖係顯示無凸緣成形測試之開始前之平面之模式 圖。 第7C圖係顯示無凸緣成形測試之結束後之縱截面之模 式圖。 10 200900981 第8A圖係顯示使用於成形解析之元素分割較小者之模 式圖。 第8B圖係顯示使用於成形解析之元素分割較大者之模 式圖。 • 5 第9圖係顯示元素大及元素小時之最大主應變分布之 解析結果之特性圖。 第10圖係顯示個人使用者終端裝置之内部構造之模式 圖。 ( 【實施方式3 10 用以實施發明之最佳形態 本發明人已先就解析對象零件之破壞部位之變形狀態 詳細加以調查。結果，得知實際發生破裂之處表現了峰值， 該峰值之附近，板厚減少率及應變量等變形量則漸次減 少。即，可推論變形集中於解析對象零件之預定領域（元素) 15 後，進而於該領域内發生變形之局部化，最後導致破壞發 生。換言之，其意味解析對象零件之破壞部位之所謂變形 ί 梯度較大。所謂變形梯度，係指解析對象零件之某部位之 依板厚減少率及應變量等變形量之位置不同之變化量（梯 度）。變形梯度係以位置（距離）微分變形量所得之微分係 20 數，使用於微小領域時，可以諸如變形梯度=[變形量/距離 (mm)]表示之。 - 藉有限元素法進行解析對象零件之變形解析而判別破 壞之習知方法，一般係採用比較依計算求得之各領域（分割 成網目狀之各元素）之變形量及另行求出之材料之破壞限 11 200900981 之方法即，習知方法(使用有限元素法之變形解析），係 於某元素之變形量超過依據解析對象零件之材料之破壞限 度而規定之破壞限度時，判定該部位為破壞部位。 然而此時將產生以下之各種問題。 10 15 20 、，有限凡素法細就各元素計算之變形量作為該元素内 ^均值。因此’若將元素大小設定較大，則存在有變形 ^大元素中’前述部位將局部存在於該元素内 *圍内此日^，即便該部㈣局部超過破壞限度， 元細已平均化變形量，亦即該部位之變形量已為平 破壞^收’而可能使作為元素内之平均之輸出值未超過 々又。此時’則無法判^該部位為破壞部位。 之元^考=應變歡局部化，而產生加以分割成甚小 之極大影變有^素法之處理時間受元素大小與總元素數 素，則右加以分割成可對應變形之局部化之甚小元 與元素之將㈣極長之相。具體而言，處理時間 元素大小若二咸=之3次方之倒數成比例。舉例言之， 使用一邊2 /2,則處理時間約為8倍，若！/4則約為64倍。 元素時，舉例t夕 之規模而言，·Φ1Λ, ± 之’以―般騎對象零件 而使用-邊〇/ Γ日禮度之處理時間，若為提昇精度 處理時間，而素則需要約64倍之640小時程度之 而將降低實用性。 小於求取解批大小右較小’亦將發生以下問題。即，使用 解析對象零件之材料之破壞限度時之標距（測量 12 200900981 大小時，將 此時，即需 破壞部之應變時作為基準之標點間距離）之元素 無法直接比較來自元素之輸出值與破壞限度。 要進行若干修正。20 • The destruction of the swearing... 々 珂 珂 珂 珂 珂 珂 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第 第The first step, the second step, and the third step are described in the field of the first division. Further, in the first step of the first step, the damage analysis method of any one of the first and second fields is divided into the end portions of the first field and the second field, respectively. Forming analysis is performed. ▲7, using the finite element method, (4) split the field of money into the field of transformation, and then form the analysis; the second step 'calculate the maximum principal strain or plate thickness reduction rate for each of the above fields; the third step 'combined abutment In the above two or more fields, the maximum principal strain or the thickness reduction rate of the aforementioned field is calculated; and, in the fourth step, the difference between the maximum principal strain or the thickness reduction ratio before and after the combination of the above-mentioned fields is selected. The aforementioned field in which the value is larger than the predetermined value is a damage-destroying portion of the component to be analyzed. 6. The method for predicting damage according to 5. In the eighth step, R is in the above-mentioned field 7 200900981, and the end portion of the analysis target component is divided, and then the forming analysis is performed. 7. The arithmetic processing device, the damage prediction method for the analysis target component, includes the "11-mechanism" using the finite element method, and the analysis target component is divided into the second field of the first field and the second field of the first field. The second mechanism calculates the maximum principal strain or the thickness reduction rate for each of the second field and each of the second fields; and the third mechanism corresponds to the analysis target component. At the position of the portion, the aforementioned maximum principal strain or the reduction in the thickness of the plate is selected as described above in which the difference between the i-th field and the second field is greater than a predetermined value! The field is a dangerous part of the damage of the aforementioned analysis target part. 8. If the calculation of 7_ is processed, the above! The mechanism determines the size of the first field and the size of the second field in accordance with the relationship between the components and the components described above. 2 9. The 胄 arithmetic processing device, the destruction of the material of the object to be resolved, the pre-emptive method package 3 has. Institutions, using the finite element method, will analyze the image of the image, and divide the "number field, and then form the analysis; % organization, the maximum main strain or plate thickness reduction rate in the front block area; the magical mechanism, the knot eight ^ 2 More than one of the above, calculate the combined field: the major principal strain or the aforementioned plate thickness reduction rate; and, the fourth mechanism, select the above, ^ the difference between the maximum principal strain or the plate thickness reduction ratio of d is greater than The IT phase domain of the devaluation value, as a pre-processing analysis (4) of the damage hazard section 10. The computer can perform the following steps: the second step uses the finite element method to use the J-field and the larger than the mussel A. The second field is greater than the first field in the first field. 20 200900981. The analysis target component is divided and the shaping analysis is performed. In the second step, the maximum principal strain or the plate thickness reduction rate is calculated for each of the first domain and the second domain; And in the third step, the difference between the first principal strain and the plate thickness reduction rate is selected at a position corresponding to the same portion of the analysis target component; and the difference between the first domain and the second domain is greater than a predetermined value. The first mentioned above Domain, as the analysis target part of the damage risk site. 11. The program according to 10., in the first step, determining a size of the first field and a size of the second field f according to a relationship between an η value of the analysis target component. \10 12. The program of 10. or 11. In the third step, if the first field having the predetermined value is not selected, at least the first field is in the first field and the second field. The setting is made smaller, and the first step, the second step, and the third step are performed in sequence. 13. The program according to any one of 10 to 12, wherein in the first step, the first field and the second field are divided into the end portions of the analysis target parts, and the forming analysis is performed. I 14. A program that causes a computer to perform the following steps: In the first step, the finite element method is used to divide the analytic object into a complex field and then perform shape analysis; and in the second step, the maximum principal strain is calculated for each of the aforementioned fields. 20 plate thickness reduction rate; in the third step, combining the above-mentioned two or more adjacent fields, calculating the aforementioned maximum principal strain or the aforementioned plate thickness reduction-rate in the above-mentioned field; and, in the fourth step, selecting the combination of the aforementioned fields The aforementioned field in which the difference between the maximum principal strain or the thickness reduction ratio of the front and the back is larger than a predetermined value is a damage danger portion of the analysis target component. 9 200900981 15. As in the program of 14., in the first step, the end portion of the analysis target component is divided by the above-described field, and the forming analysis is performed. 16. A recording medium that can be read by a computer and recorded in any of 10. to 15. 5 According to the present invention, the destruction of the machined parts can be predicted, and the dependence on the selection of the analysis conditions can be reduced, and the dangerous parts can be easily and surely selected. In this way, in addition to reducing the cost of development, lightweight materials can be applied to processed parts with higher strength. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a flow chart of a damage prediction method (apparatus) of the present invention. Fig. 2 is a characteristic diagram showing the simulation results when the upper and lower limits of the coarse and fine element division are determined. Fig. 3 is a characteristic diagram showing the simulation results when the upper and lower limits of the coarse element division are determined. 15 Fig. 4 is a characteristic diagram showing the simulation results when the upper and lower limits of the fine element division are determined. Fig. 5 is a flow chart showing the destruction prediction method (apparatus) of the present invention. Fig. 6 is an explanatory view of the shape of the plain plate used in the forming test. Fig. 7A is a diagram showing the longitudinal section of the longitudinal section before the start of the flangeless forming test. Figure 7B shows a pattern of the plane before the start of the flangeless forming test. Figure 7C shows a schematic view of the longitudinal section after the end of the flangeless forming test. 10 200900981 Figure 8A shows a model diagram for the smaller element division used for shaping analysis. Fig. 8B is a view showing a mode in which the element division for forming analysis is larger. • 5 Figure 9 is a characteristic diagram showing the analytical results of the largest principal strain distribution of the element and the hour of the element. Fig. 10 is a schematic view showing the internal structure of the personal user terminal device. (Embodiment 3 10 The best mode for carrying out the invention) The present inventors have examined in detail the state of deformation of the damaged portion of the analysis target component. As a result, it was found that the peak where the actual fracture occurred is expressed, and the peak is near The amount of deformation such as the thickness reduction rate and the strain amount is gradually reduced. That is, it can be inferred that the deformation is concentrated on the predetermined area (element) 15 of the analysis target part, and then the localization of the deformation occurs in the field, and finally the destruction occurs. In other words, it means that the so-called deformation ί gradient of the damaged portion of the analysis target part is large. The deformation gradient refers to the amount of change in the position of the deformation amount of the part to be analyzed and the position of the deformation amount such as the strain amount (gradient) The deformation gradient is the number of differential systems 20 obtained by the position (distance) differential deformation. When used in a small field, it can be expressed as a deformation gradient = [deformation amount / distance (mm)] - Analysis by finite element method The conventional method of discriminating the deformation of the object part and discriminating the damage is generally based on the comparison of the various fields obtained by calculation (divided into The amount of deformation of each element of the target and the damage limit of the material obtained separately. 11 The method of 200900981 is that the conventional method (deformation analysis using the finite element method) is based on the deformation of an element exceeding the part to be analyzed. When the damage limit of the material is specified and the damage limit is specified, the part is determined to be a damaged part. However, the following various problems are caused at this time. 10 15 20 , The amount of deformation calculated by each element is defined as the element within the element ^ The mean value. Therefore, if the element size is set to a large size, there is a deformation. In the large element, the above part will be locally present in the element*, and even if the part (4) exceeds the damage limit, the element is averaged. The amount of deformation, that is, the amount of deformation of the part has been flattened, and may cause the average output value within the element to not exceed 々. At this time, 'the part cannot be judged to be the damaged part. Test = Straining is localized, and the maximum distortion that is divided into very small is changed. The processing time of the method is determined by the element size and the total element number, and then the right is divided into the corresponding deformation. The very small element and the element will be (4) extremely long phase. Specifically, the processing time element size is proportional to the reciprocal of the third power = 2, for example, using one side of 2 /2, the processing time is about It is 8 times, if !/4 is about 64 times. In the case of the element, for example, the scale of the 夕 夕, Φ1Λ, ± 'uses the general purpose of the object to be used - edge / day processing time If it is to improve the precision processing time, and the factor needs about 64 times 640 hours, it will reduce the practicality. Less than the solution size is smaller than the right size. The following problem will occur. That is, the material of the analytical object part will be used. The elemental distance at the time of the damage limit (measurement 12 200900981, the time, the distance between the punctuation points at which the strain of the damaged part is required as a reference) cannot directly compare the output value and the damage limit from the element. A number of corrections are required.
進而，即便已將元素大小分割至極小，亦可能無法正 5確判定破壞發生之有無。即，即便解析對象零件中存在有 變形量巨大至已足以發生破壞之程度之部位，但^位係 包含較大範目而具有大致均-變形量之部&，亦可能因變 形未集中而不發生破壞。舉例言之，可以形成於解析對象 零件之孔部之周緣上產生大致均一之變形量之所謂毛邊變 10形為例。此時，即便實際上並未發生破壞，對應該部位之 凡素之輸出值亦超過破壞限度，而可能制定為破壞部位。 如以上之說明，藉習知之破壞判定方法，需要高产之 技能方可進行正確之破壞判定，且可能因變形之發生形能 及没定條件不同而遺漏破壞部位。 15 纟發明人為改善上述狀況，著重於破壞部位周圍之變 形梯度較大之性質，而發明了利用有限元素法之解析需要 進仃文7L素大小影響之平均化之新破壞判定方法。 本發明中，對於變形梯度部位，依有限元素法而使用 分割大小不同之2種元素(在此為求簡便，以較小者為第】元 如素’較大者則為第2元素）進行解析。#限元素法係將該元 素内之變形量平均後再加以輸出。因此，某元素内若存在 變形梯度較大之變形部位，該元素若為第1元素或第2元 素，則前者之輸出值將大於後者。 本發明中，利用第！元素與第2元素之平均值之差異， 13 200900981 在相當於解析對㈣件之同—部位之位置上，將就第！元素 及第2兀素分別進行解析。此時，若第以素與第^素之平 均值之輸出不同，則可推論該元素内存在變形梯度。上述 輸出值之差異則對應變形梯度之大小。變形梯度愈大，破 5壞之危險度愈高，並可依解析值之差之大小判別破壞危險 度。 本4月中，除如上述般使用分割大小不同之2種元素， 亦可構成就财大奴元素進行解析後，結合2個以上之元 素，而取元素結合前後之輸出值之差。此時，元素結合前 1〇後之平均值之輸出若有不同，則可推論該元素内存在變形 梯度，變形梯度愈大，破壞危險度愈高，並可依解析值之 差之大小判別破壞危險度。 此之所明解析值，係指一般破壞判定所使用之板厚、 板厚減v Φ帛大主應變、最大、最小主應變所代表之成 15形限度線圖上之成形裕度等，而可使用其中任何值。就解 析上之處理之容易度而言，宜使用板厚減少率或最大主應 變。又’成形加工之有限元素法解析，_般係使用在板厚 方向上不具節點而由板面方向之複數節點構成之平面(2次 元)元素’本發明雖適用於該元素，但對於才奉狀加工品所使 2〇用之1次兀几素(棒狀元素），更詳而言之對用於提昇板厚 方向之變形之解析精度之3次元元素（立體元素），本發明亦 可完全同樣地適用。 使用上述方法，即可簡便且確實地進行以往因受破壞 部位之局部變形之程度，及決定破壞限度時之測量方法等 14 200900981 之影響’非使元素大小最佳化’則難以進行之破壞判定。 具體而言，依據本發明，可完全解決上述之習知技術 之各種問題。 亦即，本發明無須使用習知技術之極小元素作為第】 5 元素或結合前之元素，故可使處理時間大幅縮短。又，此 時’無須使用小於求出解析對象零件之材料之破壞限度時 之標距之元素大小’故可直接比較來自元素之輸出值與破 壞限度。 亦即’本發明逆向操作有限元素法使元素内之變形量 10平均化之方式’而使用大小不同之2種元素。因此，相對於 以往使元素内之變形量平均化，亦即使變形量較大之部位 之影響為平均值所吸收’本發明則可正確判定破壞危險部 位。 又，一如毛邊變形，即便解析對象零件存在有變形量 15巨大至足以發生破壞之程度之部位，但該部位係包含較大 範圍而具有大致均一變形量之部分，即使因變形未集中而 不發生破壞，本發明亦可充分對應之。即，此時，該部位 之變形梯度較小（或幾無變形梯度），故若取第1元素與第2 元素之輸出值之差，則其值較小，而可正確判定其並非破 2〇 壞危險部位。 ’經本發明人鑽研檢討後，發現各種破壞中，就 d種稱為伸緣成形破壞之變形形態，破壞判定之精度可較 Μ更為提昇。所謂伸緣成形成形，係指可見於車身 外飾板之部分之巾柱之連接Μ部位，或構件類之烊接 15 200900981 所需之無凸緣加工部位等處，變形狀態近似一軸拉伸者。 上述變形形態下，破壞危險部位之變形梯度極大。且，變 形與其它破壞形態相較，亦更傾向局部化。因此，以一般 有限元素法進行解析時，必須使用極小之元素，除計算時 5間極長’亦難以連結計算值與以某種特定之標距測定後之 材料之破壞限度值。 10 15 相對於此’可知本發明藉改變用以算出解析值之元素 大小，即可以變形梯度作為解析值之差而加以評價，並確 實選出破壞危險部位。本發明若適用於因與變形梯度之關 係而易於維持伸緣成形破壞之位伸強度44〇Mpa級以上之 高強度鋼板，則可大幅提昇其預測精度，而較為適用。 另，本發明不限於有限元素法，凡可行元素分割之解 析方法均可適用。又，不僅限於成形時之破壞，亦可有致 應用於衝撞變形時之材料之破壞預測等。 以下，具體說明本發明。 -〜+贯％你如弟i圚所不，在將解析對象零件分割、 複數領域(元素）而依有限元素法崎成形解析時將使用f 小之元素（第1元素）與大於第丄元素之元素（第2元 ^ 兀f ,而進行成形解析（分割機構(步驟）u)，再就各元素* 出板厚減少率或最大主應變(解析機構(步师2 )，然後Ί 相虽於解析對象零件之同—部位之位置上，選出前述最^ 主應變或前述板厚減少率之第〗元素與第2元素之差值大、 作為解析對_，危險部位: 20 200900981 :此，分割機構U、解析機構12及選出機構15係作為 老如電腦之中央處理裝置(CPU)之各機能而實現。 利枝ί第1及第6圖中’實線代表必要機構或步驟，虛線 、代表非必要機構或步鄉。 5 10 15 20 纖情㈣件分料餘元素㈣機構(步 )日"，將顯不為3次元之零件形狀之數位資料(CAD資料 或形狀測定資料）、2次元之平面領域之集合。此時，零件 之角部之雜變化較大，故分财甚奴元素，以確保重 現性。又，解析端部之伸緣成形破壞時，宜藉元素分割而 使零件外躲無㈣而讀。又，藉元素分_成大小不 同之弟1兀素及第2元素時，可將解析對象零件整體均句細 分似或粗分化）’或就進行破壞判定之處加以細分化或粗分 化。就作業時數而言，前者較簡便，計算時間之縮短則以 後者較有利’可考量整體貞擔Μ當進行馨或混用。 在此，分割機構(步驟）11中，第1元素之大小及第2元素 之大小係依與解析對象轉之η值之而決定。 本心月中藉有限元素法而為元素分割，並進行解析 夺為重現對象#位之幾何形狀即諸如端部之曲率及角 之曲率半；^等’必須進行極微細之元素分割。又進而， 本發明中’改變7°素分料第1元素及第2元素之2種而進行 解析後就第1元素與第2元素取板厚減少率或最大主應變 之差時，必須就2元素分t'丨之大小（粗疏及細密）進行充分之 考里。本發明人已就粗疏與細密的元素分割大小之設定方 法加以鑽研檢討’而發現其與材料之加卫硬化特性有關。 17 200900981 瓜以拉伸測試求得之11值代表材料之加工硬化特性時，若 粗疏的元素分割之平均大小L coarse(單位mm)與細密的元 素分割之平均大小L fi n e (單位m m)滿足以下之關係 ，則可知 破壞預測精度良好。 5 f(2,n)^ L coarse^ f( 10,η) 〇) f(〇.5,n)^ L fine^ f(5,n) ⑺ 在此，n係材料之n值，ng 0.05即成立上式。n< 〇〇5 時’使用n=〇_〇5之值而求出L coarse及L fine即可。又，函數 f(L*，n)則如下。 10 f(L*,n)= L*(l-exp(0.37/n)/3200) (3) 上述（1)〜(3)亦可改寫如下。 2(1 —exp(0.37/n)/3200)^Lcoarse^l〇(i — exp(〇.37/n)/3200) (4) 〇·5(1 —exp(0.37/n)/3200)SL fineg 5(1 - exp(〇.37/n)/32〇〇) (5) 該函數fT遺n值而增大。n值較大時，變形較難局部化， 15故即便元素分割較大亦可確保破壞預測精度。另，n值較小 時’變形容易局部化’因此破壞危險部位之變形梯度較大， 若未進行極微細之元素分割’則破壞預測精度將降低，故 須加以對應而縮小元素分割之大小，依此而加以設定。 雖已預料η值極小而小於0.05時，可進行更小之元素分 20割，但因過小之元素分割將使計算時間大增而不甚妥適， 故可知使η值為0.05而設定之粗密之元素分割在目前之有 限元素法之數值解析精度範圍内並無應用上之問題。因 此’ η值為0.05以下時，可使η值為〇·〇5而設定元素分割。決 18 200900981 定粗疏及細密的元素分割之上下限時之模擬結果顯示於第 2圖，其特性圖則為第3及第4圖。 進而，為以高精度評價變形梯度，L fme2 比Lcoarse/Lfine為1.5以上，更宜為2以上。 5 其久’藉有限元素法進彳于成形解析，市售之軟體可使 用諸如PAM-STAMP、LS-DYNA等逐次解析型，或 AutoForm 'HyperForm等單步驟型者等，而進行零件整體 之成形步驟之解析’並鼻出各第1元素及各第2元素之板厚 減少率或最大主應變(解析機構(步驟）12)。板厚減少率與最 10大主應變係由有限元素法所使用之塑性應變增量之記錄算 出，而作為進行破壞判定之最終形狀下之值。成形解析則 可使用於包含擴孔加工之成形、無凸緣成形、伸展及深引 伸等任意之加壓成形、併用内壓之液壓成形、對管體作用 軸力與内壓而進行成形之液壓成形等。 15 在此，上述之板厚減少率或最大主應變之差係以元素 分割之大小最小時之解析結果為基準，而選出最接近所著 重之元素之位置之其它解析結果之元素，並計算其等之差。 其人，選出上述板厚減少率成最大主應變之差大於預 定值之元素作為破壞危險部位(遽出機構(步驟）15)。 20 在此，上述之預定值可作為戚壞限度值而藉其它測試 求出，或作為對應結合元素之大小之值而藉進行簡易形狀 零件之成形解析而求出。 具體而言，舉例言之，使用/邊2mm者作為第丨元素， 並使用一邊4mm者作為第2元素時，以變形量作為最大主應 19 200900981 變時之預定值宜為0.01〜0.50之範圍内之值。在此，若使用 小於0.01之值，可能受到數值解析之誤差之影響而發生判 定錯誤，或將變形梯度較小之部位亦辨識成破壞危險部 位，若使用大於0.50之值，則亦可能無法辨識變形梯度較 5 大之部位為破壞危險部位，故無法以高精度界定變形部 位。因此，宜使用0.01〜0.50之範圍内之值。 上述範圍中，以0.03〜0.20之範圍内之值為佳。進而， 0.05〜0.10之範圍内之值則更佳。 另，以變形量為板厚減少率時之預定值宜為0.01〜0.25 1〇 之範圍内之值。在此，若使用小於0.01之值，可能受到數 值解析之誤差之影響而發生判定錯誤，或將變形梯度較小 之部位亦辨識成破壞危險部位，若使用大於0.25之值，則 亦可能無法辨識變形梯度較大之部位為破壞危險部位，故 無法以高精度界定變形部位。因此，宜使用0.01〜0.25之範 15 圍内之值。 上述範圍中，以0.02〜0.15之範圍内之值為佳。進而， 0.025-0.10之範圍内之值則更佳。 上述之解析（解析機構（步驟）12)與選出（選出機構（步 驟）15)可於同一電腦内執行，亦可在以1電腦執行解析（解析 20 機構（步驟）12)後，再對另一電腦輸入已改變該解析結果之 元素分割之大小之2種以上之各元素之板厚減少率或最大 主應變（輸入機構（步驟）13)，再執行選出（選出機構（步 驟）15)。 2.之本發明中，參照第2〜第4圖，如上所述，在分割機 20 200900981 構（步驟）11中，第1元素之大小及第2元素之大小係依與解析 對象零件之η值之關係而決定。 3.之本發明中，在破壞危險部位選出（選出機構（步 驟）15)時，若未選出大於上述預定值之第1元素，則第1元素 ' 5 及第2元素中至少將第1元素設定成較小，再進行分割（分割 機構（步驟）11)，並算出各元素之板厚減少率或最大主應變 (解析機構（步驟）12)，再選出破壞危險部位（選出機構（步 驟）15)。 f 4.之本發明中，在第1圖之分割機構（步驟）11中，係將 10 解析對象零件之端部分割成複數元素而進行成形解析，並 於選出機構（步驟）15中選出任一端部作為破壞危險部位。 將解析對象零件之端部分割成複數元素時，係特別就 進行破壞判定之部分進行分割而確實使元素分割之大小改 變。進行破壞判定之端部無論元素分割之大小，均須無凹 15 凸而平滑地連接。又，為確實進行端部之破壞判定，評價 沿行端部之變形梯度甚為重要，元素分割之大小宜在沿行 ί 端部之方向上確實有改變（參照第8A圖及第8B圖）。 選出任一端部作為破壞危險部位時，與1.之發明相 同，係選出各預定元素之板厚減少率或最大主應變之差大 20 於預定值之元素部位作為破壞危險部位。 ' 5.之本發明如第5圖所示，係將解析對象零件分割成複 - 數元素（分割機構（步驟）21)，並依有限元素法進行成形解 析，就各元素算出板厚減少率或最大主應變（解析機構（步 驟)22)，然後結合隣接之2個以上之元素，而算出已結合之 21 200900981 ㉞^應變(算出機構(步綱，再選 ^前後之板厚減少率或最大主應變之差大於預定值之 作為破壞危險部位(選出機構(步驟)25)。 在此’分割機構21、解析機構22、算出機構24及選出 機構25可作為諸如電 书細之中央處理裝置（CPU)之各機能而 Η現。 驟、21、將解析對象零件分割成複數元素（分割機構(步 '=1)時’將顯示為3次元之零件形狀之數位資料(CAD資料 10 15 2〇 s化狀,則疋貝料）、2次元之平面領域之集合。此時，零件 =部之雜f化較A，故分料甚小之元素，以確保重 估愛i又’解析端部之伸緣成形破壞時，㈣元素分割而 使零件外周線無凹凸而平滑。 其次，使用與第！圖之解析（解析機構（步驟）12)相同之 人體’進行與1.之發_同之成形解析，而進行*件整# =成形步驟之解析，並算出所著重之各元素之板厚減少率 :最大主應變(解析機構(步驟降板厚減少率、最大主應 變之計算則與第丨圖之解析(解析機構(步驟)12)相同。 接著，結合隣接之2個以上元素時，需要結合對象之各 =素之計算值與各元素德置(絲）㈣。結錢元素之計 鼻值(板厚減少率或最大主應變）係各元素之計算值之算術 平=。結合後元素之位制可為各元素位置之算術平均， 更簡便者亦可直接沿用中央部元素之位置。 ^繼之，就元素之結合前後之板厚減少率之差比較結合 則後之後，分別選出位置最接近之元素，計算上述元素之 22 200900981 板厚減少率之差。至於最大主 最接近之元素社差。 〜變’料算結合前後位置 然後，選出上述元素之結人^^ 主應變之差大於預定值之元二後之板厚減少率或最大 預定值之計算，則與、圖之\破壞危險部位⑼。 相同。 圮出（選出機構（步驟）15) 上述之解析（解析機構（步驟)22) 驟)24)可於同一電腦内連續 ;、出(异出機構(步 (解析機構(步驟)22)後，再對另_ t 析 10 15 20 4之板厚減少率或最大主應變(輸人機構(步驟仰 仃异出（算出機構(步驟)24)、選峨出機構(步驟)25)。 6.之本發明與4•之本發明相同，係對5之本發明適用4 之本發明之構造者。 · 7·之本發明係對應1 ·之破壞預測方法之發明之演算處 理襞置之發明，於第1圖中，將步驟改為機構即可理解。处 解析機構12可安裝與L之發明所說明之市售之軟體相 同之軟體而使用。 本裝置具有可將已分割之各元素已求出之板厚減少率 或最大主應變輸入其它電腦之輸入機構13。輸入機構射 使用鍵盤、滑鼠、各種數位轉化器等。 在此，輸入機構13及選出機構15亦可與分割機構叹 解析機構12構成不同裝置。此時，以—電腦完成成开；解才 之結果作為原始資料而加以輸入至另一電腦，即可並 理而提昇效率。 23 200900981 8. 之本發明係對應2.之破壞預測方法之發明之演算處 理裝置之發明，於第1圖中，將步驟改為機構即可理解。 9. 之本發明係對應5.之破壞預測方法之發明之演算處 理裝置之發明，於第5圖中，將步驟改為機構即可理解。 5 在此，輸入機構23、算出機構24及選出機構25亦可與 分割機構21及解析機構22構成不同之裝置。此時，以一電 腦完成成形解析之結果作為原始資料而加以輸入至另一電 腦，即可並行處理而提昇效率 10. 之本發明係對應1.之破壞預測方法之電腦程式之 10 發明，即第1圖中用於實施各步驟之電腦程式。 輸入步驟13可為藉鍵盤輸入之步驟，亦可為在程式 内，將解析步驟12所算出之板厚減少率或最大主應變自動 輸入選出步驟15(讀取資料)之步驟。 11. 之本發明係對應2_之破壞預測方法之電腦程式之 15 發明，即第1圖中用於實施各步驟之電腦程式。 12. 之本發明係對應3.之破壞預測方法之電腦程式之 發明，即第1圖中用於實施各步驟之電腦程式。 13. 之本發明係對應4 ·之破壞預測方法之電腦程式之 發明，即第1圖中用於實施各步驟之電腦程式。 20 14.之本發明係對應5.之破壞預測方法之電腦程式之 發明，即第5圖中用於實施各步驟之電腦程式。 輸入步驟23可為藉鍵盤輸入之步驟，亦可為在程式 内，將解析步驟22所算出之板厚減少率或最大主應變自動 輸入算出步驟24(讀取資料)之步驟。 24 200900981 15_之本發明係對應6.之破壞預測方法之電腦。 發明’即第5圖中用於實施各步驟之電腦程式 16_之發明係一種記錄媒體，可由電腦進行讀取，記錄 有上述10.〜15.之任一項之電腦程式，即，軟碟片、CD汉等 [第1實施例] 以下藉實施例說明本發明。先將一般由圓筒衝孔器實 施之擴孔測試之素板形狀加以分割而模擬無凸緣成形之成 形實驗。即，將於180mm見方之素板之中心部設有孔洞（直 徑60mm，或40mm、20mm)者，如第6圖所示般裁成1/4，再 10如第7A〜第7C圖所示，將被加工板4以壓料板2固定於肩部 R5mm之106φ模座1上，然後使用肩部Ri〇mm之1〇〇φ圓筒 平底衝孔器3進行成形。此時，除緣高度5在孔徑60mm時約 為20mm ’ 40mm時約為30mm，20mm時約為40mm。素材則 使用板厚1.6mm之440MPa級冷軋鋼板。實驗中以4片一組進 15 行成形。其結果匯整表示於表1，孔徑60mm時，中央部雖 已發生破裂’但孔徑4〇mm及20mm時，未發生破裂，而可 實現無凸緣成形。 已進行模擬該實驗結果之有限元素法解析。領域（已分 割成網目狀之各元素）大小則為約2mm(第8A圖）與約 20 4mm(第8B圖）之2種而準備了業經元素分割之素板。分割則 使用藉CAD而製作之形狀資料，在指定圓周部之元素分割 數後’藉電腦自動進行分割。 此外之解析條件，則兩者皆相同。成形解析係藉 PA Μ - S ΤΑ Μ P而進行。由解析後之整體之資料就分割後之各 25 200900981 元素選出成形後之最大主應變與板厚值，並由成形後之板 厚算出板厚減少率為（初始板厚一成形後板厚）/(初始板 厚）。所得之值則與圓周部上之各元素之位置資訊一同輸 出，而輸入資料解析用之電腦。 5 第9圖係已輸入資料解析用電腦之最大主應變之資 料，就元素較小（約2mm)與元素較大（約4mm)時個別顯示之 特性圖。如該圖所示，可知元素較小時，最大主應變之最 大值較大，且分布圖斜度較大。此可推論其代表在該條件 下，在圓周之中央部產生較大之變形梯度。元素較小時之 10 最大主應變之最大值之元素之位置與其絕對值則已先求 出。然後，在資料解析用電腦内界定最接近元素較大時之 計算結果中表現為元素較小時之最大值之元素之位置，而 求出該最大主應變之絕對值。最後則於電腦上計算該2個絕 對值之差。上述操作等於取第9圖之元素大與元素小之結果 15 之峰值之差。 其結果顯示於表1。又，以相同方式求出之板厚減少率 之差亦已顯示在同一表内。孔徑60mm時差較大，相對於 此，孔徑愈小，差亦減小。差較大代表變形梯度較大，並 對應了實驗中孔徑60mm時發生破裂。本實施例中，破壞部 20 位因伸緣成形變形而形成一軸拉伸狀態，等向同性材料之 板厚減少率則為最大主應變之約1/2。因此，解析判定值可 採用任一種，但為突顯差異，宜採用絕對值較大之最大主 應變。本實施例中，用以判定破壞危險部位之預定值之解 析值之差之絕對值雖因使用之元素大小而改變，而難以特 26 200900981 定之，但推論此次之檢討範圍内，最大主應變可為0.05程 度，板厚減少率則可為0.025程度。 另，本實施例所判定之破壞預測部位以第8A圖之A點 代表之。 5【表1】 孔徑 (mm) 成形測試結果 元素大與元素小時之 解ϋ值之j: 解析之 判定 測試與 解析之 對應 最大主應變 板厚減少率 60 NG :中央部 發生破裂 0.074 0.039 NG ◎ 40 OK :可成形 0.029 0.017 OK ◎ 20 OK :可成形 0.012 0.017 OK ◎ [第2實施例] 本例使用第1實施例中孔徑60mm、元素較小（約2mm) 之解析結果，而藉結合隣接之2個以上元素，並比較結合前 10 後之差，而評價變形梯度，以調查是否可判定破壞。 元素分割及成形解析已進行與第1實施例之元素較小 時相同之處理（第8A圖） 先由成形解析結果輸出元素（尤其以解析值為峰值之 元素附近為中心）之解析值與其位置資訊。對資料解析用電 15 腦輸入該資料，而就此次選出之隣接之2〜5個結合元素個別 算出解析值之算術平均，並算出與先前之解析之解析值之 最大值之差。 由隣接之2元素平均後之最大主應變之分布求出之最 大值與平均前之最大值之差為0.007，與三元素之平均值之 27 200900981 差為〇·02，與4元素之平均值之差為0.035，與五元素之平均 值之差為〇._。與帛1實施例所示之實際上改變元素大小 而冲算者崎’則為較小值，但可知藉計算依隣接之複數 凡素結合後之元素算出之解析值與結合前之解析值之差， 5即可知變形梯度之大小，即，可選出破壞危險部位。此時， 變开/梯度之大小與結合後元素之大小之比，則可決定應結 合幾個元素，但宜取複數之結合元素數，並調查解析值之 差之影響度。本實施例中，係取4元素結合之平均值與結合 刖之解析值之差，可知將最大主應變〇〇3程度以上設定成 1〇發生破壞之預定值，即可判定破壞。 [第3實施例] 本例在第1實施例之孔徑40mm之測試條件下，調查是 否可預測各種強度之材料之破壞。使用材料係表2所示之軟 鋼至_MPa級之鋼板。使用材料板厚為i 6_。 15 實驗之結果，98GMpa級鋼板於無凸緣部位之中央部發 生了伸緣成形破裂。在實驗之相同條件下亦進行了有限元 素法解析元素大小為約2mm與約4咖之2種而進行解析。 ㈣之7?㈣如第8ASJ及第8B圖所示，端部無凹凸而平滑 地連接’且確實注意沿端部而改變元素大小，故指定圓周 部之分割數後，即由電腦進行自動分割。成形解析、各元 素之最大主應變及板厚減少率之計算則與第！實施例相同。 —在恥件τ輸出成形解析結果後，即與第1實施例 以資料解析用電腦計算無凸緣成形後之最大主應變 及板厚減少率之最大值之差。 其結果則顯示於表2。材料強 28 200900981 度愈高’差亦愈大’可知變形集中部位之變形梯度較大。 與第1實施例相同，將最大主應變之G.05之差以上判定為破 壞，則就98〇偷級鋼板判定為破壞，可知與實驗結果-致。 【表2】 成形測試 結果 οκ ： 可成形 OK ： 可成形 OK ： 可成形 NG ： 中央部 發生破裂 ----— 元素大與元素小時 值之差 解析之 測試與 解析之 對應 最大 主應變 — 板厚 減少率 判定 0.018 ---— 0.008 OK ◎ 0.029 ----- 0.017 OK ◎ 0.035 --- 0.021 OK ◎ 0.068 0.031 NG ◎ 材料 鋼種 降伏強度 (MPa) 拉伸強度 (MPa) 延展率 (%) 軟鋼 190 326 43 440 MPa級 295 449 36 590 MPa級 340 612 33 980 MPa級 752 1034 15 (應用本發明之其它實施例) 上述本實施例之破壞預測方法（第i圖之分割步驟n〜 勒步驟15，及第5圖之分割步驟21〜選出步驟糊可藉儲 10 存於電腦之讀姐轉之程•作而實現。絲式及記 錄有該程式之可供電腦讀取之記錄媒體包含於本發明之 内。 體而。刖牡式可^錄於諸如cd_rqM等記錄媒 =或經由各種傳輸媒體，而提供電腦使用。可記錄前述 =之轉媒體除CD-ROM以外，亦可使用軟碟片 、硬碟、 媒二Γ片、非揮發性記憶卡等。另，前述程式之傳輸 =則可使用以程式資訊作為载波而傳送供給之電腦網路 統之通訊媒體。在此，所謂電腦網路係指LAN、網際網 29 200900981 路等WAN、無線通訊網路等，通訊媒體則指光纖電纜等有 線電路及無線電路等。 又’本發明所包含之程式不僅限於以電腦執行傳入之 程式而實現上述實施例之機能者。舉例言之，該程式與電 5腦内動作之〇s(作業系統）或其它應用程式軟體等共同實2 上述實施例之機能時之此種程式，亦包含在本發明之内。 又，所傳入之程式之處理之全部或一部分藉電腦之功能擴 充板或功能擴充單元而執行實現上述實施例之機能時之此 種程式，亦包含在本發明之内。 10 舉例言之，第10圖係顯示個人使用者終端裝置之内部 構造之模式圖。該第10圖中，1200係設有CPU1201之個人 電腦(PC)°PC1200可執行儲存於ROM1202或硬碟(HD) 1211 内，或由軟碟機(FD) 1212傳入之裝置控制軟體。該pci2〇〇 可全面控制與系統匯流排1204連接之各裝置。 15 PC1200 之 CPU120 卜 ROM1202 或硬碟(HD) 1211 中儲存 之程式’則可實現本實施例之第1圖之分割步驟u〜選出步 驟15 ’及第5圖之分割步驟21〜選出步驟25之步驟等。 1203係RAM，作為CPU1201之主記憶體、工作區等而 作用。1205係鍵盤控制器（KBC)，用以控制來自鍵盤（KB) 20 1209及未圖示之裝置等之指示輸入。 1206係CRT控制器（CRTC)，用以控制CRT顯示器（CRT) 1210之顯示。1207係磁碟控制器（DKC)。DKC1207則控制 對儲存開機程式、複數之應用程式、編輯資料夾、使用者 資料夾、以及網路管理程式等之硬碟（HD)1211及軟碟機(FD) 30 200900981 1212之存取。在此，所謂開機程式，係指啓動程式：用以 開始電腦之硬體及軟體之執行(動作）之程式。 1208係網路介面卡(NIC)，可經由LAN1220而與網路印 表機、其它網路機器或其它PC進行雙向資料交換。 5 產業之可利用性 依據本發明而預測加工零件之破壞，可降低對解析條 件之選擇之依賴性，而輕易且確實地選出破壞危險部位。 藉此，除可降低開發所需之成本，亦可對加工零件應用強 度更高之材料，而實現輕量化。 10 【圖式簡單說明】 第1圖係本發明之破壞預測方法（裝置）之流程圖。 第2圖係顯示粗疏及細密的元素分割之上下限決定時 之模擬結果之特性圖。 第3圖係顯示粗疏的元素分割之上下限決定時之模擬 15 結果之特性圖。 第4圖係顯示細密的元素分割之上下限決定時之模擬 結果之特性圖。 第5圖係本發明之破壞預測方法（裝置）之流程圖。 第6圖係成形測試所使用之素板形狀之說明圖。 20 第7A圖係顯示無凸緣成形測試之開始前之縱截面之模 式圖。 第7B圖係顯示無凸緣成形測試之開始前之平面之模式 圖。 第7C圖係顯示無凸緣成形測試之結束後之縱截面之模 31 200900981 式圖。 第u圖係顯示使用於成形解析之元素分害 式圖。 第8B圖係顯示使用於成形解析之元素分割較大者之模 式圖。 第9圖係顯示元素大及元素小時之最大主應變分布之 解析結果之特性圖。 第10圖係顯示個人使用者終端裝置之内部構造之模式 圖。 ° 【主要元件符號說明】 1…模座 1200.·.個人電腦(pC) 2…壓料板 1202 …ROM 3···圓筒平底衝孔器 1203 …RAM 4···被加工板 1204…系統匯流排 5···除緣高度 1205…鍵盤控制器(KBC) “···分割機構丨步驟） 1206· · · CRT控制器(CRTC) 12···解析機構(步驟） 1207".磁碟控制器(DKC) 13···輸入機構(步驟） 1208…網路介面卡(NIC) 15…選出機構(步驟） 1209…鍵盤(KB) 21·.·分割機構(步驟） 1210…CRT顯示器(CRT) 22···解析機構(步驟） 1211…硬碟(HD) 23···輸入機構(步驟） 1212…軟碟機(FD) 24.·.算出機構(步驟） 1220··-LAN 25…選出機構(步驟） R…肩部 32Furthermore, even if the element size has been divided to a minimum, it may not be possible to determine the presence or absence of the damage. That is, even if there is a portion in the analysis target part where the amount of deformation is large enough to cause damage, the portion having a larger range and having a substantially uniform-deformation amount may also be due to the fact that the deformation is not concentrated. No damage occurs. For example, a so-called burr-deformed shape which is formed on the periphery of the hole portion of the analysis target part to produce a substantially uniform deformation amount can be exemplified. At this time, even if the damage does not occur in fact, the output value of the corresponding part exceeds the damage limit, and may be determined as a damaged part. As explained above, by using the method of determining the damage of the knowledge, it is necessary to have a high-yield skill to perform the correct damage determination, and the damage may be missed due to the deformation of the shape and the undetermined condition. 15 In order to improve the above situation, the inventors focused on the nature of the deformation gradient around the damage site, and invented the new damage determination method that requires the averaging of the influence of the size of the 7L element by the analysis of the finite element method. In the present invention, for the deformation gradient portion, two elements having different division sizes are used according to the finite element method (here, for the sake of simplicity, the smaller one is the first element), and the larger one is the second element. Analysis. The #limit element method averages the amount of deformation within the element and outputs it. Therefore, if there is a deformation part with a large deformation gradient in an element, if the element is the first element or the second element, the output value of the former will be larger than the latter. In the present invention, the use of the first! The difference between the average of the element and the second element, 13 200900981 will be the same as the position of the same part of the analysis of the (four) pieces! The element and the second element are separately analyzed. At this time, if the output of the average value of the first element and the second element is different, it can be inferred that there is a deformation gradient in the element. The difference in the above output values corresponds to the magnitude of the deformation gradient. The larger the deformation gradient, the higher the risk of breaking 5, and the damage risk can be judged by the difference of the analytical values. In the middle of this month, in addition to the above two types of elements having different split sizes, it is also possible to combine the two or more elements in the analysis of the elements of the rich and the slave, and take the difference between the output values before and after the combination of the elements. At this time, if the output of the average value of the element before the first one is different, it can be inferred that there is a deformation gradient in the element, the larger the deformation gradient, the higher the damage degree, and the damage can be determined according to the difference of the analytical values. Risk. The known analytical value refers to the thickness of the plate used for the general damage determination, the thickness of the plate minus v Φ 帛 the main strain, the forming margin on the 15-shaped limit line graph represented by the maximum and minimum principal strains, and the like. Any of these values can be used. In terms of the ease of processing on the analysis, it is preferable to use the plate thickness reduction rate or the maximum main strain. In addition, the finite element method analysis of the forming process uses a plane (2 dimensional element) composed of a plurality of nodes in the direction of the plate thickness without a node in the direction of the plate thickness. The present invention is applicable to the element, but The third processing element (rod element) used for the processing of the product, and more specifically, the ternary element (stereo element) for improving the resolution of the deformation in the thickness direction, the present invention may also It applies exactly the same. By using the above method, the degree of local deformation due to the damaged portion and the measurement method when determining the damage limit can be easily and surely performed. 14 The impact of the 200900981 effect is not difficult to determine the damage of the element size. . In particular, according to the present invention, various problems of the above-described conventional techniques can be completely solved. That is, the present invention does not require the use of a very small element of the prior art as the element 5 or the element before the combination, so that the processing time can be greatly shortened. Further, at this time, it is not necessary to use an element size smaller than the gauge length when determining the damage limit of the material of the analysis target part, so that the output value and the damage limit of the element can be directly compared. That is, the 'reverse operation finite element method of the present invention averaging the deformation amount 10 in the element' uses two kinds of elements having different sizes. Therefore, even if the amount of deformation in the element is averaged in the past, even if the influence of the portion having a large deformation amount is absorbed by the average value, the present invention can correctly determine the damage end portion. Further, as the burr is deformed, even if the part to be analyzed has a portion where the amount of deformation 15 is large enough to cause damage, the portion contains a large range and has a portion having a substantially uniform deformation amount, even if the deformation is not concentrated. The damage may occur, and the present invention may also fully correspond. That is, at this time, the deformation gradient of the portion is small (or a few deformation gradients), so if the difference between the output values of the first element and the second element is taken, the value is small, and it can be correctly determined that it is not broken. Destroy dangerous parts. After reviewing the results of the inventors' research, it was found that in various kinds of damages, the deformation type of the d-form is called the deformation of the edge forming, and the accuracy of the damage determination can be improved. The so-called stretch forming means the joint portion of the towel column which is visible in the outer panel of the vehicle body, or the flangeless processing portion required for the joint of the member type 200900981, and the deformation state is approximately the one-axis stretcher. . In the above deformation mode, the deformation gradient of the damaged portion is extremely large. Moreover, the deformation is more localized than other failure modes. Therefore, when parsing by the general finite element method, it is necessary to use a very small element, and it is difficult to connect the calculated value with the damage limit value of the material measured by a certain gauge length, except for the calculation of the five extremely long lengths. 10 15 With respect to this, it can be seen that the present invention evaluates the element size for calculating the analytical value, that is, the deformation gradient can be evaluated as the difference between the analytical values, and the damage-destroying portion is surely selected. When the present invention is applied to a high-strength steel sheet having a tensile strength of 44 〇 Mpa or more which is easy to maintain the deformation of the edge forming deformation due to the relationship with the deformation gradient, the prediction accuracy can be greatly improved, and it is suitable. Further, the present invention is not limited to the finite element method, and any analytical method of feasible element division can be applied. Further, it is not limited to the damage at the time of molding, and may be applied to the prediction of damage of the material at the time of collision deformation. Hereinafter, the present invention will be specifically described. -~+%% If you are analysing object parts, complex fields (elements) and finite element method, you will use f element (1st element) and greater than 丄 element. The element (the second element ^ 兀f , and the forming analysis (dividing mechanism (step) u), and then the thickness reduction rate or the maximum principal strain of each element * (analytical mechanism (step 2), and then In the position of the same part of the analysis target part, the difference between the first element main strain or the reduction factor of the plate thickness and the second element is selected as the analysis pair _, the dangerous part: 20 200900981: The division mechanism U, the analysis mechanism 12, and the selection mechanism 15 are realized as functions of a central processing unit (CPU) of a computer such as a computer. In the first and sixth figures, the solid line represents a necessary mechanism or step, and a dotted line represents Non-essential institutions or step-by-step. 5 10 15 20 Fibrous (four) parts of the remaining elements (four) mechanism (step) day ", will be not the three-dimensional part of the shape of the digital data (CAD data or shape determination data), 2 The collection of the planar fields of the dimension. At this point, the corner of the part Miscellaneous changes are large, so the wealth is a slave element to ensure reproducibility. Also, when the analysis of the extension of the end is broken, it is better to use the element to separate the parts to avoid reading (4). In the case of a child of different sizes, the second element and the second element may be subdivided or roughly differentiated in the overall sentence of the object to be analyzed, or subdivided or roughly differentiated in the case of damage determination. The former is simpler, and the shortening of the calculation time is more advantageous for the latter. It can be considered as a whole or a mixed use. Here, in the segmentation mechanism (step) 11, the size of the first element and the size of the second element are dependent. It is determined by the η value of the analytic object. In the heart of the month, the finite element method is used to divide the element, and the resolution is taken as the object of reproduction. The geometric shape of the bit is the curvature of the end and the curvature of the corner half; In the present invention, it is necessary to perform the extremely fine element division. In the present invention, the amount of the first element and the second element is reduced by the change of the first element and the second element of the 7° element. Or the difference between the maximum principal strain, it must be 2 yuan The size of the t'丨 (sparse and fine) is fully tested. The inventors have conducted a review on the setting method of the coarse and fine element size, and found that it is related to the hardening characteristics of the material. 17 200900981 When the 11 value obtained by the tensile test represents the work hardening property of the material, the average size L coarse (unit mm) of the coarse element division and the average size L fi ne (unit mm) of the fine element division satisfy the following Relationship, then we can see that the prediction accuracy of the damage is good. 5 f(2,n)^ L coarse^ f( 10,η) 〇) f(〇.5,n)^ L fine^ f(5,n) (7) Here, The n value of the n-based material, ng 0.05, holds the above formula. n < 〇〇 5 o' Use the value of n = 〇 _ 〇 5 to obtain L coarse and L fine. Also, the function f(L*,n) is as follows. 10 f(L*,n)= L*(l-exp(0.37/n)/3200) (3) The above (1) to (3) can also be rewritten as follows. 2(1 —exp(0.37/n)/3200)^Lcoarse^l〇(i — exp(〇.37/n)/3200) (4) 〇·5(1 —exp(0.37/n)/3200) SL fineg 5(1 - exp(〇.37/n)/32〇〇) (5) This function fT increases the value of n. When the value of n is large, the deformation is difficult to localize, so that even if the element is divided large, the prediction accuracy can be ensured. In addition, when the value of n is small, the deformation is easily localized, so the deformation gradient of the damaged part is large. If the elemental division is not performed, the prediction accuracy of the damage will be reduced, so the size of the element division must be reduced accordingly. Set accordingly. Although it is expected that the value of η is extremely small and less than 0.05, a smaller elemental division can be performed. However, since the elemental division is too small, the calculation time is greatly increased, which is not so proper. Therefore, it is known that the η value is 0.05 and the coarseness is set. The elemental segmentation has no application problems within the range of numerical resolution accuracy of the current finite element method. Therefore, when the η value is 0.05 or less, the element division can be set by setting the η value to 〇·〇5.决 18 200900981 The simulation result of the upper and lower limits of the coarse and fine element division is shown in Fig. 2, and the characteristic diagram is the 3rd and 4th. Further, in order to evaluate the deformation gradient with high precision, L fme2 is 1.5 or more, more preferably 2 or more, than Lcoarse/Lfine. 5 For a long time, the finite element method is used for forming analysis. The commercially available software can be formed by using a analytic type such as PAM-STAMP or LS-DYNA, or a single-step type such as AutoForm 'HyperForm. In the analysis of the steps, the plate thickness reduction rate or the maximum principal strain of each of the first element and each of the second elements is extracted (analysis means (step) 12). The plate thickness reduction rate and the maximum principal strain system are calculated from the record of the plastic strain increment used by the finite element method, and are used as the value under the final shape for the damage determination. The forming analysis can be used for hydraulic forming including any forming such as reaming, flangeless forming, stretching, and deep drawing, hydroforming by internal pressure, and axial force and internal pressure acting on the tube. Forming, etc. 15 Here, the difference between the plate thickness reduction rate and the maximum principal strain is based on the analysis result when the size of the element division is the smallest, and the element of the other analysis result closest to the position of the element to be emphasized is selected and calculated. Wait for the difference. As a person, the above-mentioned element whose plate thickness reduction rate is the maximum main strain difference larger than a predetermined value is selected as the damage danger portion (the ejection mechanism (step) 15). Here, the predetermined value described above can be obtained by other tests as a corruption limit value, or can be obtained by performing forming analysis of a simple shape component as a value corresponding to the size of the bonding element. Specifically, for example, when a side of 2 mm is used as the second element and a side of 4 mm is used as the second element, the amount of deformation is the maximum main requirement. 19 200900981 The predetermined value of the change is preferably 0.01 to 0.50. The value inside. Here, if a value smaller than 0.01 is used, a judgment error may occur due to an error in numerical analysis, or a portion having a small deformation gradient may be recognized as a damaged portion. If a value larger than 0.50 is used, it may not be recognized. The part with a deformation gradient of 5 is a dangerous part, so the deformation part cannot be defined with high precision. Therefore, it is preferred to use a value in the range of 0.01 to 0.50. Among the above ranges, values in the range of 0.03 to 0.20 are preferred. Further, a value in the range of 0.05 to 0.10 is more preferable. Further, the predetermined value when the amount of deformation is the thickness reduction ratio is preferably a value in the range of 0.01 to 0.25 1 Torr. Here, if a value smaller than 0.01 is used, a judgment error may occur due to an error in numerical analysis, or a portion having a small deformation gradient may be recognized as a damaged portion. If a value larger than 0.25 is used, it may not be recognized. The part with a large deformation gradient is a dangerous part, so the deformation part cannot be defined with high precision. Therefore, it is preferable to use a value within a range of 0.01 to 0.25. Among the above ranges, values in the range of 0.02 to 0.15 are preferred. Further, values in the range of 0.025 to 0.10 are more preferable. The above analysis (analysis mechanism (step) 12) and selection (selection mechanism (step) 15) can be performed in the same computer, or after performing analysis by one computer (analysis 20 mechanism (step) 12), and then The computer selects the plate thickness reduction rate or the maximum principal strain (input means (step) 13) of each of the two or more elements whose size is divided by the analysis result, and performs the selection (selection means (step) 15). 2. In the present invention, referring to the second to fourth figures, as described above, in the splitter 20 200900981 configuration (step) 11, the size of the first element and the size of the second element depend on the η of the analysis target part. Determined by the relationship of values. 3. In the present invention, when the destruction target portion is selected (the selection mechanism (step) 15), if the first element larger than the predetermined value is not selected, at least the first element is selected from the first element '5 and the second element. When it is set to be small, it is divided (dividing mechanism (step) 11), and the plate thickness reduction rate or the maximum main strain of each element (analysis mechanism (step) 12) is calculated, and the damage dangerous part is selected (selection mechanism (step) 15). In the present invention, in the dividing mechanism (step) 11 of Fig. 1, the end portion of the 10 analysis target component is divided into a plurality of elements, and the forming analysis is performed, and the selecting means (step) 15 is selected. One end serves as a dangerous part. When the end portion of the analysis target component is divided into a plurality of elements, the portion where the damage determination is performed is divided, and the size of the element division is surely changed. The end portion where the damage determination is performed, regardless of the size of the element division, must be smoothly connected without the concave and convex portions. Further, in order to surely determine the damage of the end portion, it is important to evaluate the deformation gradient along the end portion of the line, and the size of the element division should be changed in the direction along the end of the line ί (refer to Figs. 8A and 8B). . When either end portion is selected as the damage-destroying portion, the same as the invention of 1. The element portion where the difference in the thickness of the predetermined element or the maximum principal strain is greater than a predetermined value is selected as the damage-destroying portion. According to the fifth aspect of the present invention, the analysis target component is divided into complex-number elements (dividing mechanism (step) 21), and the forming analysis is performed by the finite element method, and the plate thickness reduction rate is calculated for each element. Or the maximum principal strain (analytical mechanism (step) 22), and then combine the two or more adjacent elements to calculate the combined 21 200900981 34 ^ strain (calculation mechanism (step, re-selection before and after the thickness reduction rate or maximum The difference between the main strains is greater than the predetermined value as the damage danger portion (selection mechanism (step) 25). Here, the 'dividing mechanism 21, the analysis mechanism 22, the calculation mechanism 24, and the selection mechanism 25 can be used as a central processing device such as an electric book ( Steps 21, the object to be resolved is divided into complex elements (when the segmentation mechanism (step '=1)' will be displayed as a 3-dimensional part shape digital data (CAD data 10 15 2〇) s morphological, then mussel material), the collection of the planar field of 2 dimensions. At this time, the part = part of the impurity is more than A, so the material is very small, to ensure that the re-evaluation of love and 'analysis end When the extension edge is broken, (4) the element is divided The outer circumference of the piece is smooth without unevenness. Next, using the same human body as the analysis of the figure (analysis mechanism (step) 12), the forming analysis is performed in the same manner as in the case of 1., and the forming step is performed. The analysis and calculation of the plate thickness reduction rate of each element to be emphasized: maximum principal strain (analysis mechanism (step reduction of plate thickness reduction, calculation of maximum principal strain is the same as analysis of the second figure (analysis mechanism (step) 12) Next, when combining two or more adjacent elements, it is necessary to combine the calculated values of the respective elements of the object with the respective elements (wire) (4). The nose value (plate thickness reduction rate or maximum principal strain) of the money-making element is The arithmetic value of the calculated value of each element =. The bit system of the combined element can be the arithmetic mean of the position of each element, and the simpler one can directly follow the position of the central element. ^ Following, the thickness of the element before and after the combination of the elements After comparing the difference between the reduction rates, the elements with the closest position are selected, and the difference between the 22 200900981 plate thickness reduction ratios of the above elements is calculated. As for the largest and closest elemental difference, the change is calculated. , the calculation of the knot of the above-mentioned elements ^^ The difference between the main strain and the maximum predetermined value after the second value of the predetermined value is greater than the predetermined value, and the same as the damage to the dangerous part (9). (Step) 15) The above analysis (analysis mechanism (step) 22)) 24) can be continued in the same computer; (out of the mechanism (step (analysis mechanism (step) 22), then another _ t Analysis of the plate thickness reduction rate or maximum principal strain of 10 15 20 4 (input mechanism (step calculations (step 24), selection mechanism (step) 25). 6. The invention and 4 The present invention is the same as the constructor of the present invention applicable to the present invention of 5. The present invention is the invention of the arithmetic processing of the invention corresponding to the destruction prediction method of Fig. 1. In Fig. 1, the steps can be understood by changing the steps to the mechanism. The analysis unit 12 can be installed by using the same software as the commercially available software described in the invention of L. The apparatus has an input mechanism 13 for inputting the plate thickness reduction rate or the maximum principal strain which has been obtained for each of the divided elements into another computer. The input mechanism uses a keyboard, a mouse, various digital converters, and the like. Here, the input mechanism 13 and the selection mechanism 15 may be configured differently from the division mechanism saliency analysis mechanism 12. At this point, the computer is used to complete the opening; the result of the solution is input as input to another computer, and the efficiency can be improved. 23 200900981 8. The present invention is an invention of an arithmetic processing apparatus corresponding to the invention of the damage prediction method of 2. In the first figure, the step is changed to a mechanism. 9. The present invention is an invention of an arithmetic processing apparatus corresponding to the invention of the damage prediction method of 5. In Fig. 5, it is understood that the steps are changed to a mechanism. Here, the input mechanism 23, the calculation mechanism 24, and the selection mechanism 25 may be configured differently from the division mechanism 21 and the analysis mechanism 22. In this case, the result of the forming analysis performed by one computer is input as input data to another computer, and the processing can be performed in parallel to improve the efficiency. 10. The present invention relates to a computer program corresponding to the destruction prediction method of 1. The computer program used to implement each step in Figure 1. The input step 13 may be a step of inputting by means of a keyboard, or may be a step of automatically inputting the plate thickness reduction rate or the maximum principal strain calculated in the analysis step 12 into the step 15 (reading data) in the program. 11. The present invention is a computer program corresponding to the computer program of the destruction prediction method of 2, which is a computer program for implementing each step in Fig. 1. 12. The present invention is an invention of a computer program corresponding to the destruction prediction method of 3. The computer program for implementing each step in Fig. 1. 13. The present invention is an invention of a computer program corresponding to the destruction prediction method of Fig. 4, that is, a computer program for implementing each step in Fig. 1. The invention of the computer program corresponding to the damage prediction method of 5. is the computer program for carrying out the steps in Fig. 5. The input step 23 may be a step of inputting by means of a keyboard, or may be a step of automatically inputting the plate thickness reduction rate or the maximum principal strain calculated in the analysis step 22 into the step 24 (reading data) in the program. 24 200900981 15_ The invention is a computer corresponding to the destruction prediction method of 6. The invention of the computer program 16_ for performing the steps in the fifth embodiment is a recording medium which can be read by a computer and recorded with the computer program of any of the above 10.~15. Sheet, CD Han, etc. [First Embodiment] The present invention will be described below by way of examples. The shape of the plain plate test, which was generally performed by a cylindrical punch, was first divided to simulate a forming experiment without flange forming. That is, a hole (60 mm in diameter, or 40 mm, 20 mm) will be provided in the center of the 180 mm square plate, as shown in Fig. 6, cut into 1/4, and then 10 as shown in Figs. 7A to 7C. The machined plate 4 is fixed to the 106φ die holder 1 of the shoulder R5mm by the press plate 2, and then formed by using the 1〇〇φ cylinder flat bottom punch 3 of the shoulder Ri〇mm. At this time, the edge height 5 is about 30 mm at a hole diameter of 60 mm and about 40 mm at 20 mm, and about 40 mm at 20 mm. For the material, a 440 MPa grade cold-rolled steel sheet having a thickness of 1.6 mm was used. In the experiment, 14 rows were formed in groups of 4 pieces. The results are shown in Table 1. When the hole diameter is 60 mm, the center portion has cracked. However, when the hole diameter is 4 mm and 20 mm, no cracking occurs, and flangeless molding can be realized. A finite element method analysis that simulates the results of this experiment has been performed. The size of each of the fields (the elements which have been divided into meshes) is about 2 mm (Fig. 8A) and about 20 4 mm (Fig. 8B), and the element plate divided by the element is prepared. In the case of segmentation, the shape data created by CAD is used, and after the number of elements in the specified circumference is divided, the computer automatically divides. In addition, the analytical conditions are the same. The forming analysis is performed by PA Μ - S ΤΑ Μ P. From the analysis of the overall data, each of the 25 200900981 elements is divided into the maximum principal strain and the plate thickness after forming, and the plate thickness reduction rate is calculated from the formed plate thickness (initial plate thickness - formed plate thickness) / (Initial board thickness). The value obtained is output together with the position information of each element on the circumference, and the computer for inputting data is analyzed. 5 Figure 9 is the characteristic data of the maximum principal strain of the computer for input data analysis, which is displayed separately for small elements (about 2 mm) and large elements (about 4 mm). As shown in the figure, it can be seen that when the element is small, the maximum value of the maximum principal strain is large, and the slope of the distribution map is large. It can be inferred that under this condition, a large deformation gradient is generated in the central portion of the circumference. The position of the element with the maximum value of the 10 largest principal strain when the element is small and its absolute value are first obtained. Then, in the computer for data analysis, the position of the element whose maximum value is the largest when the element is the largest is defined, and the absolute value of the maximum principal strain is obtained. Finally, the difference between the two absolute values is calculated on the computer. The above operation is equivalent to taking the difference between the peak of the element of Fig. 9 and the result of the small element 15 . The results are shown in Table 1. Further, the difference in the thickness reduction ratios obtained in the same manner is also shown in the same table. The time difference of the aperture of 60 mm is large, and as a result, the smaller the aperture is, the smaller the difference is. The larger difference indicates that the deformation gradient is larger and corresponds to the crack at 60 mm in the experiment. In the present embodiment, the fracture portion 20 is formed in a one-axis tension state due to the deformation of the extension edge, and the plate thickness reduction rate of the isotropic material is about 1/2 of the maximum principal strain. Therefore, any of the analytical judgment values may be used, but in order to highlight the difference, it is preferable to use the maximum principal strain having a large absolute value. In this embodiment, the absolute value of the difference between the analytical values of the predetermined values for determining the damage to the dangerous part is changed by the size of the element used, and it is difficult to determine the maximum principal strain within the scope of the review. It can be 0.05 degrees, and the plate thickness reduction rate can be about 0.025. Further, the damage prediction portion determined in the present embodiment is represented by point A in Fig. 8A. 5 [Table 1] Aperture (mm) Forming test result element size and element hour solution value j: Analytical judgment test and analysis corresponding maximum main strain plate thickness reduction rate 60 NG: Central part cracking 0.074 0.039 NG ◎ 40 OK : Formable 0.029 0.017 OK ◎ 20 OK : Formable 0.012 0.017 OK ◎ [Second Embodiment] This example uses the analysis result of the aperture of 60 mm and the small element (about 2 mm) in the first embodiment. The two or more elements are compared with the difference between the first and last 10, and the deformation gradient is evaluated to investigate whether the damage can be determined. The element division and the shape analysis have been performed in the same manner as when the element of the first embodiment is small (Fig. 8A). The analysis value and the position of the output element (especially the vicinity of the element whose resolution value is the peak value) are first formed by the shaping analysis result. News. For data analysis, the brain inputs the data, and the arithmetic mean of the analytical values is calculated for each of the adjacent 2 to 5 combined elements selected this time, and the difference from the maximum value of the previously analyzed analytical values is calculated. The difference between the maximum value obtained by the distribution of the maximum principal strains of the adjacent two elements and the maximum value before the average is 0.007, and the difference between the average of the three elements of 27 200900981 is 〇·02, and the average of the four elements. The difference is 0.035, and the difference from the average of the five elements is 〇._. The actual value of the element is changed as shown in the embodiment of 帛1, and the calculator is smaller. However, it can be understood that the analytical value calculated by the combination of the elements of the adjacent plural elements and the analytical value before the combination are calculated. Poor, 5 can know the size of the deformation gradient, that is, the damage can be selected. At this time, the ratio of the size of the open/gradient to the size of the combined element determines that several elements should be combined, but it is preferable to take the number of complex elements and investigate the influence of the difference in the analytical value. In the present embodiment, the difference between the average value of the combination of the four elements and the analytical value of the combined enthalpy is determined, and it is understood that the maximum principal strain 〇〇3 or more is set to a predetermined value at which the rupture occurs, and the damage can be determined. [Third embodiment] In the present example, under the test conditions of the aperture of 40 mm of the first embodiment, it was investigated whether or not the destruction of materials of various strengths was predictable. The materials used are the mild steels shown in Table 2 and the steel plates of the _MPa grade. The thickness of the material used is i 6_. 15 As a result of the experiment, the 98GMpa grade steel plate broke out at the center of the flangeless portion. Under the same conditions of the experiment, the finite element method was also used to analyze the element size of about 2 mm and about 4 coffees for analysis. (4) 7? (4) As shown in Figures 8ASJ and 8B, the ends are smoothly connected without unevenness and 'it is really important to change the size of the elements along the ends. Therefore, after dividing the number of divisions of the circumference, the computer automatically divides . The forming analysis, the calculation of the maximum principal strain and the thickness reduction rate of each element are the same as the first! The examples are the same. - After the shavings τ output forming analysis results, the difference between the maximum principal strain and the maximum thickness reduction rate after flangeless forming is calculated by the computer for data analysis in the first embodiment. The results are shown in Table 2. Material strength 28 200900981 The higher the degree, the larger the difference is. The deformation gradient of the concentrated part of the deformation is large. In the same manner as in the first embodiment, when the difference between the maximum principal strain and the G.05 was judged to be broken, the 98-inch stealth steel sheet was judged to be broken, and it was found that the results were in agreement with the experimental results. [Table 2] Forming test result οκ : Formable OK : Formable OK : Formable NG : Crack in the center ----- The difference between the element size and the element hour value is the maximum principal strain of the test and analysis. Thickness reduction rate judgment 0.018 --- 0.008 positive ◎ 0.029 ----- 0.017 OK ◎ 0.035 --- 0.021 OK ◎ 0.068 0.031 NG ◎ Material steel grade drop strength (MPa) Tensile strength (MPa) Extension rate (%) Mild steel 190 326 43 440 MPa grade 295 449 36 590 MPa grade 340 612 33 980 MPa grade 752 1034 15 (other embodiments of the invention are applied) The above-described damage prediction method of the present embodiment (the division step n to the step of the i-th diagram) 15, and the segmentation step 21 of the fifth figure - the selection step paste can be realized by the process of storing and storing the computer in the computer. The silk type and the recording medium for recording the computer for reading the program are included in The invention can be used for recording on a recording medium such as cd_rqM or via various transmission media, and can be used for computer use. The above-mentioned media can be recorded in addition to the CD-ROM, and a floppy disk can also be used. Slice, hard drive Media two-chip, non-volatile memory card, etc. In addition, the transmission of the above program can use the program information as a carrier to transmit the communication medium supplied to the computer network. Here, the so-called computer network refers to the LAN, Internet 29 200900981 WAN, wireless communication network, etc., communication media refers to fiber-optic cables and other wired circuits and wireless circuits, etc. Further, the program included in the present invention is not limited to the implementation of the above-mentioned embodiments by executing an incoming program by a computer. For example, the program is also included in the present invention when the program is compatible with the functions of the above-described embodiments, such as the operation of the brain (operation system) or other application software. Moreover, such a program when all or part of the processing of the incoming program is executed by the function expansion board or the function expansion unit of the computer to implement the functions of the above embodiments is also included in the present invention. 10 For example, Fig. 10 is a schematic view showing the internal structure of a personal user terminal device. In Fig. 10, a 1200 is equipped with a personal computer (PC) of the CPU 1201. The PC 1200 can be stored in the R. Device control software introduced in the OM1202 or hard disk (HD) 1211, or from the floppy disk drive (FD) 1212. The pci2〇〇 can fully control the devices connected to the system bus 1204. 15 CPU120 of the PC1200 Bu ROM1202 or The program stored in the hard disk (HD) 1211 can implement the steps from the dividing step u to the selecting step 15' of the first embodiment of the present embodiment and the dividing step 21 to the selecting step 25 of the fifth drawing. The 1203 series RAM functions as a main memory, a work area, and the like of the CPU 1201. The 1205 Series Keyboard Controller (KBC) is used to control the input of inputs from the keyboard (KB) 20 1209 and devices not shown. The 1206 Series CRT Controller (CRTC) is used to control the display of the CRT Display (CRT) 1210. 1207 is a disk controller (DKC). The DKC1207 controls access to the hard disk (HD) 1211 and floppy disk drive (FD) 30 200900981 1212 for storing boot programs, multiple applications, editing folders, user folders, and network management programs. Here, the "boot program" refers to the startup program: a program for starting the execution (action) of the hardware and software of the computer. The 1208 is a network interface card (NIC) that can exchange bidirectional data with a network printer, other network machine, or other PC via the LAN1220. 5 INDUSTRIAL APPLICABILITY According to the present invention, it is possible to predict the destruction of a machined part, thereby reducing the dependency on the selection of the analysis condition, and easily and surely selecting a damage-damaged part. In this way, in addition to reducing the cost of development, it is possible to apply a stronger material to the machined part and achieve weight reduction. 10 [Simplified Schematic Description] Fig. 1 is a flow chart showing a method (device) for destroying the present invention. Fig. 2 is a characteristic diagram showing the simulation results when the upper and lower limits of the coarse and fine element division are determined. Fig. 3 is a characteristic diagram showing the results of the simulation 15 when the upper and lower limits of the coarse element division are determined. Fig. 4 is a characteristic diagram showing the simulation results when the upper and lower limits of the fine element division are determined. Fig. 5 is a flow chart showing the destruction prediction method (apparatus) of the present invention. Fig. 6 is an explanatory view of the shape of the plain plate used in the forming test. 20 Figure 7A shows a schematic view of the longitudinal section before the start of the flangeless forming test. Fig. 7B is a pattern diagram showing the plane before the start of the flangeless forming test. Figure 7C shows the mode of the longitudinal section after the end of the flangeless forming test. 31 200900981 Equation. Figure u shows the elemental damage diagram used for forming analysis. Fig. 8B is a view showing a mode in which the element division for forming analysis is larger. Fig. 9 is a characteristic diagram showing the analytical results of the maximum principal strain distribution of the element large and element hours. Fig. 10 is a schematic view showing the internal structure of the personal user terminal device. ° [Main component symbol description] 1... Mold base 1200.·. Personal computer (pC) 2...Blank plate 1202 ...ROM 3···Cylinder flat bottom punch 1203 ...RAM 4···Processed plate 1204... System bus 5···Edge height 1205...Keyboard controller (KBC) “···Split mechanism 丨Steps] 1206· · · CRT controller (CRTC) 12··· Analytical mechanism (step) 1207".Magnetism Disc Controller (DKC) 13···Input Mechanism (Step) 1208...Network Interface Card (NIC) 15...Selection Mechanism (Step) 1209...Keyboard (KB) 21···Split Mechanism (Step) 1210...CRT Display (CRT) 22··· Analytical mechanism (step) 1211... Hard disk (HD) 23···Input mechanism (step) 1212... floppy disk drive (FD) 24.·. Calculation mechanism (step) 1220··-LAN 25...selection mechanism (step) R...shoulder 32